v0.5.80
v0.5.80New Features
- Convert spoken commands for "slash" and "new line" into their respective characters for improved transcription accuracy.
- Introduce an explicit spoken language setting, allowing users to specify their preferred language for transcription.
- Implement quick-prompt hotkeys on Windows for faster access to common commands.
- Enable no-speech detection hints for macOS users, helping to avoid incorrect transcriptions when the microphone is active.
- Preserve URLs during voice translations to maintain formatting and context.
Improvements
- Mute system output while dictating on macOS and Windows, providing a clearer dictation experience.
- Enhance the dictionary functionality on Windows with fuzzy and phonetic matching to improve word recognition.
- Improve the transcription process by blocking no-speech Whisper hallucinations, ensuring more accurate results.
- Show detected source language in reader mode, enhancing user awareness of the current language context.
- Provide session-aware autostart on system reboot, ensuring the app is ready for use immediately after startup.
Bug Fixes
- Resolve issues with excessive blank lines during text insertion, leading to cleaner transcriptions.
- Fix the "Check for Updates" notification to appear in all app states, ensuring users are informed about available updates.
- Address problems with the tray menu on macOS, ensuring it displays correctly on secondary monitors.
- Correct the exact-match functionality for short words in the dictionary to reduce false positives.
- Prevent crashes during silent-engine recovery and improve overall stability during use.